|Publication number||US7272067 B1|
|Application number||US 11/060,925|
|Publication date||Sep 18, 2007|
|Filing date||Feb 18, 2005|
|Priority date||Aug 22, 2003|
|Also published as||US6897543|
|Publication number||060925, 11060925, US 7272067 B1, US 7272067B1, US-B1-7272067, US7272067 B1, US7272067B1|
|Inventors||Cheng H. Huang, Yowjuang Liu, Chih-Ching Shih, Hugh Sung-Ki O|
|Original Assignee||Altera Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (22), Non-Patent Citations (13), Referenced by (33), Classifications (30), Legal Events (2)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application is a division of patent application Ser. No. 10/646,013, filed Aug. 22, 2003 now U.S. Pat. No. 6,897,543, which is hereby incorporated by reference herein in its entirety.
This invention relates to antifuses, and more particularly, to electrically-programmable integrated circuit antifuses.
Programmable fuses and antifuses are used in a variety of integrated circuit applications. For example, a programmable logic device may have logic that is customized by programming appropriate fuses or antifuses on the device. Fuses and antifuses may also be used to permanently switch redundant circuitry into place to fix reparable defects during the integrated circuit manufacturing process. Sometimes it may be desired to use fuses or antifuses to program a serial number or other special information into a circuit (coding).
Fuses and antifuses may be programmed using special laser-based systems or may programmed electrically.
Laser-programmable fuses are often used for integrated circuits that have aluminum interconnects. Polysilicon and aluminum fuses can be blown open by focusing a precisely-aligned laser beam on the appropriate fuses. Because each fuse must be serially-programmed, the programming process can be lengthy. The programming equipment used in laser-based systems may also be complex and expensive.
There are additional concerns when using laser-based fuses with the increasingly-popular copper interconnects. Because copper has a high reflectivity and low coefficient of thermal expansion, it is generally more difficult to use a laser to blow open a copper fuse than an aluminum fuse. There is a range of acceptable laser energies that can be used to properly blow a copper fuse. The laser energy must be sufficient to complete the laser cut through the copper. At the same time, the laser energy cannot be too high to avoid cracking.
Laser-based antifuses have also been proposed. With this type of arrangement, a vertical conducting link between two adjoining metal layers may be formed by application of the laser beam. As with laser-based fuses, laser antifuses only work within a certain range of laser energies. Laser-based antifuse programming equipment can also be complex and expensive.
Electrically programmable fuses are attractive because the need for laser-based programming equipment is eliminated and programming speeds can generally be increased.
Non-volatile memory fuses such as fuses based on electrically-programmable read-only memory (EPROM) technology or electrically-erasable programmable read-only memory (EEPROM) technology have been used. These fuses generally require high voltages to store charges in their floating gate memory cells. The high voltages may be supplied using external equipment or using integrated charge pump circuitry. While it is possible to scale down the programming voltages of these floating-gate cells by thinning the tunnel dielectric, the tunnel dielectric cannot be too thin or the data retention time of the cell will suffer. Moreover, it can be difficult to integrate non-volatile memory-based fuses into many logic circuit designs, because the special fabrication steps needed to produce the non-volatile memory-based fuses may add otherwise unnecessary constraints to the logic fabrication process.
Another type of electrically-programmable fuse that has been investigated uses silicided polysilicon devices. When high current is applied to these fuses, the resulting agglomeration of salicide on top of the polysilicon resistor increases its resistance. The programming of this type of fuse typically requires application of a programming current of 20 milliamps for a duration of 100 ms. The sensing circuits used with such fuses also tend to dissipate a large amount of power when the fuses are in their unprogrammed states. The programming process also may not be entirely permanent, because over time cobalt may migrate into the boundaries between silicide grains, thereby reducing resistance.
Electrically-programmed antifuses have been proposed that are programmed using insulator breakdown. A thin insulator is provided between two adjacent metal layers. When a sufficient voltage is applied to the metal layers, the insulator breaks down, thereby shorting the metal layers together. Although these antifuses can generally be programmed using lower powers than those used to program silicided polysilicon devices, the antifuses can be difficult to fabricate. In particular, deposition and etch-back steps needed to provide a thin oxide for the antifuses under the normally clear contact regions on the device may add undesirable complexity to the fabrication process.
Electrically-programmed antifuses based on floating-gate structures with two shortable polysilicon layers have also been proposed. With this arrangement, an oxide between the lower of the two polysilicon layers and a substrate can be broken down using a high voltage programming signal. These antifuses require 20 volts for programming. Moreover, because two polysilicon layers are required, these antifuses cannot be used on circuits that are fabricated using a single-layer polysilicon process.
It is therefore an object of the present invention to provide improved integrated circuit fuses or antifuses.
Integrated circuit antifuse circuitry is provided. Antifuses may be based on metal-oxide-semiconductor (MOS) transistors formed on the semiconductor substrate of an integrated circuit. Programming and sensing circuitry may be connected to the antifuse transistor. The programming circuitry may be used to apply appropriate programming signals to the antifuse transistor during programming. The sensing circuitry may be used to convert the resistance value of the antifuse transistor into a corresponding output voltage signal for other circuitry on the integrated circuit.
When an antifuse transistor is in its unprogrammed state, the transistor is off and its drain-to-source resistance is high. During programming, sufficient current flows through the transistor to induce localized melting in the semiconductor substrate in the vicinity of the transistor. The melted semiconductor reduces the drain-to-source resistance of a programmed antifuse transistor by a factor of about four orders of magnitude relative to an unprogrammed antifuse transistor.
The antifuse transistor has source, drain, gate, and substrate terminals. To turn on the transistor on with sufficient strength to produce melting of the semiconductor, the substrate voltage is taken about 0.7 to 1.0 volts higher than the source region voltage. This forward biases the substrate-source junction and injects carriers into the transistor in the substrate region under the gate. The carrier injection in this region turns the transistor on and produces a high current between the drain and source terminals.
The substrate voltage may be raised using any suitable arrangement. One way in which to raise the substrate voltage is using an external resistor connected between the substrate and a source of ground potential. By causing current to flow through the resistor, the substrate voltage can be raised during programming.
Current can be made to flow through the resistor using any suitable circuitry. With one suitable arrangement, a programming voltage is applied to the drain of the antifuse transistor. This causes avalanche breakdown in the drain-substrate p-n junction. The resulting avalanche current flows through the resistor and raises the substrate voltage. As the substrate voltage rises, the substrate-source junction becomes forward biased, which injects more carriers into the depletion region at the drain-substrate junction and serves as a source of positive feedback. When the carrier injection into the substrate region is sufficient, the drain-source current rises high enough to melt the semiconductor between the drain and source, thereby creating a conductive pathway that permanently reduces the antifuse transistor's resistance.
With another suitable arrangement, a number of Zener diodes (e.g., two Zener diodes) are connected between the drain and substrate. When the programming voltage is applied to the Zener diodes, it induces Zener breakdown. The current flowing through the Zener diodes during Zener breakdown flows through the resistor and bias the substrate sufficiently to turn on the antifuse transistor and melt the semiconductor. The Zener breakdown voltage of the Zener diodes may be low enough to allow programming with readily available power supply voltages such as the standard I/O power supply voltage used to operate peripheral circuitry surrounding lower-level core logic.
The substrate region lying beneath the antifuse transistor may be electrically isolated from the remaining substrate using a deep n-well isolation region.
Further features of the invention, its nature and various advantages will be more apparent from the accompanying drawings and the following detailed description of the preferred embodiments.
Programmable fuses and antifuses may be used to program (configure) portions of an integrated circuit. The integrated circuit may be, for example, a programmable logic device that has user-customizable logic. The integrated circuit may also be a microprocessor, a digital signal processor, an application specific integrated circuit, logic, memory, a digital circuit, an analog circuit with programmable devices, a combination of such circuits or any other suitable integrated circuit.
The integrated circuit may be programmed by a manufacturer or by an intermediate or end user. For example, if a manufacturer desires to include a serial number or special coding in the integrated circuit, the manufacturer can program the fuses or antifuses on the integrated circuit appropriately during the manufacturing process. As another example, a manufacturer may produce integrated circuits that contain redundant circuitry. If a repairable defect is detected by the manufacturer during device testing, an appropriate spare circuit may be switched into use to replace the defective circuit.
Programmable logic devices can be customized by users to perform desired logic functions. Such devices may be customized by programming fuses and/or antifuses appropriately. Programmable logic devices may also use fuses and antifuses to implement redundancy schemes or to store codes, serial numbers, or other special information in a non-volatile manner as with other integrated circuits.
Electric programming of fuses and antifuses is often preferable to laser-based programming, because costly and complex laser-based programming systems are not needed. Fuses conduct current when unprogrammed and become open circuits when placed in a programmed state. Antifuses are open circuited when unprogrammed and become conducting when programmed.
The present invention relates to improved antifuses.
A conventional antifuse 10 that uses an oxide-breakdown arrangement is shown in
A thin layer of insulator 26 is formed between the aluminum 14 of the B terminal and the conducting n+ silicon portion 22 of the A terminal. When the antifuse is unprogrammed, the insulator prevents current from flowing between the A and B terminals.
When it is desired to program the antifuse of
The conventional antifuse 10 of
Another conventional antifuse arrangement is shown in
In its unprogrammed state, the bottom polysilicon layer 32 of antifuse 28 is electrically isolated from p-type silicon substrate by insulating layer 37. When programming, the programming voltage for the floating gate serves to break down the insulator 37 to form conductive paths 38. The conductive paths 38 electrically connect terminals A and B.
The conventional antifuse structure of
An integrated circuit 42 in accordance with the present invention is shown in
The particular voltages described herein are merely illustrative. Any other suitable voltage levels may be used if desired. Moreover the polarities of the signals (positive and negative) and types of semiconductor (e.g., n-type and p-type silicon) that are used are merely illustrative.
Although described in the context of a silicon-based arrangement, integrated circuit 42 may be based on any suitable semiconductor. For example, the substrate material from which circuit 42 is formed may be a semiconductor such as silicon (or silicon-on-insulator), silicon germanium, gallium arsenide, indium phosphide, combinations of such materials, etc. These substrate materials are typically processed while in the form of wafers that are subsequently divided into individual circuits 42 prior to packaging.
The circuitry of integrated circuit 42 is interconnected by a number of layers of circuit interconnections. These interconnect layers include patterned conductors and insulators. The conductors of the regular metal interconnects on circuit 42 serve to route signals between various devices on circuit 42. The insulators serve to electrically isolate the conductors from each other and from the devices formed in the substrate. The layers of insulators and conductors in the interconnects and the antifuse structures, active devices, and associated circuitry of circuit 42 may be formed using plasma deposition and etching, chemical vapor deposition, sputtering and evaporation, electroplating, ion-implantation and diffusion, thermal oxide growth, and other suitable semiconductor fabrication techniques.
In a typical modern integrated circuit there are about 4-9 interconnect layers and other layers of material such as polysilicon, silicides, nitride, solder bumps, and passivation layers. For clarity, only the most relevant portions of the antifuses and associated structures are shown in the FIGS. The programming signals and power supply signals from I/O terminals such as terminals 44, 46, and 48 may be provided to circuitry within the integrated circuit such as control circuitry 50. Control circuitry 50 may increase the voltage of the power supply (if desired) and may produce programming signals at appropriate voltage and current levels for antifuse programming. As an example, control circuitry 50 may increase the strength of low-level programming signals provided at I/O pins 48 so that the resulting strengthened signals may be used to program antifuses 52.
The programmable antifuses may have associated sensing circuitry 52. The sensing circuitry may be used to convert the resistance of an antifuse (which may be either high or low depending on the state of the antifuse) into a corresponding voltage signal (e.g., a high or low logic voltage level VOUT). Each antifuse may have an associated sensing circuit. The output voltage VOUT from each antifuse sensing circuit may be applied to a corresponding controllable device (e.g., a transistor gate or gates, one or more transistor gates in a multiplexer, switch, pass transistor, or the input or control terminal of any other suitable logic element or circuit component on circuit 42). Such configurable components are depicted generally as configurable circuitry 54 in
If desired, the outputs VOUT of the sensing circuits may be read without using them to specifically configure any downstream logic. For example, if it is desired to program a serial number or other code into the circuit 42, the appropriate bits may be programmed into circuit 42 by programming antifuses accordingly. The output signals from the sensing circuitry associated with the antifuses will then represent the states of the programmed antifuses. If access to the information programmed into the antifuses is desired, circuitry 54 may be used to access and read the states of the sensing circuit outputs.
An illustrative antifuse circuit 56 for an integrated circuit such as integrated circuit 42 of
In circuit 56, resistors 60 and 62 (having respective illustrative resistances of about 5 MΩ and 10 MΩ) serve as part of a programming and sensing circuit. The sensing circuit serves to convert the programmed or unprogrammed resistance value of transistor 58 into a corresponding logic high or low signal VOUT at sensing circuit output terminal 72 during post-programming operation of circuit 42.
A power supply voltage (e.g., a voltage VCC of 1.2 volts) may be provided to terminals 64 and 66. A source of ground potential (e.g., a voltage VSS of 0 volts) may be applied to terminals 68 and 70.
A charge pump circuit 74 or other suitable circuitry may be used to generate a programming supply voltage (e.g., a voltage VPP of 5.5 volts) from voltages VCC and VSS. The programming supply voltage may be greater than the normal power supply voltage VCC used on chip 42. The use of charge pump 74 to increase the voltage level of VCC to that of VPP avoids the need to provide VPP from an external source (e.g., one of pins 48 of
The programming supply voltage VPP is used as an input to gating transistor (pass transistor) 78. The gate of transistor 78 may be controlled by appropriate control signals VCS supplied at terminal 82. With one suitable arrangement, low-level programming signals (e.g., logic-level programming information VLLCS) may be provided at terminal 80 of control circuit 76. The logic-level programming information VLLCS may, for example, be provided from external electrical programming equipment that is connected to circuit 42 through one or more of pins 48 during the configuration or programming of circuit 42 at a manufacturer as part of the manufacturing process or during the programming of circuitry 42 by an intermediate user or end user.
The control circuit 76 may use the elevated programming power supply voltage VPP and the lower-level programming information signals VLLCS to create corresponding programming control signals VCS. These control signals may be provided to the control gate of pass transistor 78. During programming, the output of pass transistor 78 (programming signal VP) is provided to output terminal 84. A typical programming signal VP has a magnitude VPP of 5.5 volts minus the threshold-voltage of the pass-gate transistor T2 and a duration of about 20-50 ms. The Vp signal can be kept the same while reducing the Vpp voltage if a PMOS-pass gate transistor is used in place of the illustrative NMOS-pass gate transistor shown in
Only a single 20-50 ms programming pulse is generally required to program each antifuse bit, so programming all of the antifuses on circuit 42 is relatively rapid. Moreover, because the power required to program each antifuse is relatively low, multiple bits may be programmed in parallel by control circuitry 50 (
When antifuse transistor 58 is unprogrammed (its initial state), the drain-to-source resistance of transistor 58 (the resistance between drain D at antifuse terminal A and source S at antifuse terminal B in
When antifuse transistor 58 is programmed by momentarily applying voltage VPP across its drain and source terminals, the drain-to-source resistance of transistor 58 drops significantly (e.g., by about four or five orders of magnitude or more). The resulting low resistance of transistor 58 pulls the voltage on terminal 72 low, so that the output voltage VOUT is at a logic low signal level.
An illustrative cross-section of a suitable device structure 86 suitable for implementing the antifuse circuitry of
Transistor 58 has a gate (labeled G in
The gate silicide materials are electrically connected to the overlying portion of metal 92 (i.e., the gate metal). Unlike the prior art arrangement of
Resistor 60 is formed from a layer of oxide 102. This oxide may be, for example the oxide layer that is referred to as the “poly-layer dielectric” in many NMOS fabrication processes and that is used to isolate poly interconnects from the substrate. A layer of shallow trench isolation (STI) 104 may be formed under oxide 102 in the region of resistor 60 to help isolate resistor 60 from substrate 88. The electrically active portion of resistor 60 may be formed from a lightly-doped (or undoped) polysilicon layer 100. Metal region 105 may contact p+ tap region 109, to ohmically connect one end of resistor 60 to the substrate 88 (shown as VSUB in
The metal regions 90, 92, 105, and 111 may be electrically isolated from the other layers of antifuse structure 86 by a patterned insulating layer 112. Layer 112 may be, for example, the insulating layer often referred to as “metal 1 dielectric” where metal 1 is the bottom layer of metal interconnect. Layer 112 may be formed from any suitable insulating material such as phosphosilicate glass (PSG), chemical vapor deposition (CVD) oxide, or other insulating dielectric films.
The details of the construction of transistors 58 and 60 that are shown in
The p-type silicon substrate 88 lies between the source and drain and is normally not conducting due to the two back-to-back diodes formed by the n+/p/n+ structure of the transistor 58. The gate is at ground, which turns the transistor off. Because the transistor is not conducting, the resistance of transistor 58 in its unprogrammed state is high. The resistance of transistor 58 may, for example, be more than 100 MΩ.
When it is desired to program antifuse transistor 58, a programming voltage may be applied to the transistor's drain terminal. For example, logic-level control signals VLLCS (e.g., a 20-50 ms positive voltage pulse) may be applied to control circuitry 76 via terminal 80 (
Taking the drain voltage to VPP produces a reverse bias in the drain-substrate junction of transistor 56, because the substrate voltage VSUB is initially grounded at VSS through resistor R2. The depletion region boundary for the drain-substrate junction is shown by dotted line 114 in
The magnitude of VPP may be significantly higher than the voltage VCC normally used to operate the transistors on circuit 56. As a result, the reverse-biased diode formed at the drain-substrate junction starts to experience avalanche breakdown. The current versus voltage characteristic of the drain-substrate junction is shown in
During avalanche breakdown, electron-hole pairs are created at the drain-substrate junction (drain junction) and current flows from the transistor's drain to the transistor's substrate. The substrate is electrically connected to ground through p+ tap region 109 and resistor 60, so the avalanche current produces a voltage drop across resistor 60 equal to the avalanche current times the resistance R2.
The depletion region's electric field E accelerates electrons 116 towards region 94 and holes 118 towards the substrate 88 and source 96. As these carriers travel through the transistor, additional carriers are freed by the avalanche breakdown effect. In addition, the voltage drop across R2 causes the substrate voltage VSUB to rise. As the substrate voltage VSUB reaches about 0.7 to 1 V, the substrate-source p-n junction becomes forward biased. This in turn injects large numbers of carriers into the substrate, as shown in
The injected carriers diffuse to the avalanche-breakdown region at the drain-substrate junction and are accelerated by the electric field at the junction. The additional injected carriers contribute to the avalanche breakdown of the junction and serve as a source of positive feedback. In effect, the source, substrate, and drain of the transistor act as the emitter, base, and collector of a parasitic bipolar transistor, so carriers injected into the substrate (base) from the source-substrate junction serve to turn the transistor on. As the process progresses, the drain-source current increases so much that an irreversible conductive path 120 is melted through the silicon between drain and source. This permanently programs the antifuse to be in a conductive “on” state.
The time required to program antifuse transistor 58 is typically 10s of milliseconds. With one suitable approach, each antifuse is programmed using a programming pulse of a fixed known duration (e.g., a pulse with a fixed duration in the range of 20-50 ms). Another suitable approach involves monitoring the status of transistor 58 during programming, so that the programming voltage may be reduced as soon as programming of the transistor 58 is complete. This approach can reduce the power dissipated during the programming process.
One way in which to monitor the status of transistor 58 is to use a feedback path such as path 122 of
If desired, the antifuse circuit may be based on reverse-biased Zener diodes. An illustrative antifuse circuit 56′ that is based on two Zener diodes 124 is shown in
Although other numbers of Zener diodes could be used (e.g., one or three or more), two Zener diodes are generally preferred. Using only one Zener diode might produce more leakage current than desired during sensing (for unprogrammed devices), which would increase power dissipation. If three Zener diodes were to be used, the required programming voltage might be higher than desired.
With the two-Zener-diode arrangement of
In general, a lower programming voltage supply level VPP′ may be used when compared to the avalanche-breakdown arrangement of
With this approach it is not necessary to provide additional charge pump circuitry 74 on circuit 42 to raise the programming voltage (although a charge pump may be used if desired to provide a higher programming voltage). Moreover, special external voltage sources and special high-voltage routing circuitry on circuit 42 need not be provided. Rather, the I/O power supply voltage that is already available (for use in operating the I/O circuitry during normal operation) can be reused for antifuse programming.
As shown in
In the antifuse circuit arrangement of
When the programming voltage VP′ is applied to the drain terminal (e.g., a voltage signal having a magnitude VPP of about 3.3 V), the reverse bias voltage across Zener diodes 124 (−3.3 volts) exceeds their combined Zener breakdown voltage (about 2 V). As a result, the diodes enter Zener breakdown and conduct a substantial amount of current (see
The current through the Zener diodes passes through the resistor R2, which leads to a corresponding voltage rise in the substrate 88 (which is electrically connected to the upper terminal of resistor 60 through tap 109). As the substrate voltage VSUB rises to about 0.7-1 volt, the source (the “base” of the parasitic bipolar transistor), remains grounded at VSS. This causes the substrate-source junction (the “base-emitter” junction of the parasitic bipolar transistor) to become forward biased.
The forward-biased substrate-source junction injects a large number of carriers (electrons) into the substrate. The substrate acts as the base in the parasitic bipolar, so when carriers are injected into the substrate, the drain-source current (the “collector-emitter” current in the parasitic bipolar) increases substantially as with any bipolar transistor. The resulting current is substantial enough to melt the silicon and thereby program the antifuse transistor.
An illustrative Zener diode structure that may be used for diodes 124 is shown in
The p-n junction of the Zener diode may be formed using p+ region 130 and n+ region 132. A layer of dielectric (e.g., the metal 1 dielectric layer of PSG) 134 may be used to define contact holes for metal layer 136 (e.g., the first metal layer M1) in the dielectric interconnect stack on the integrated circuit 42). The structure of
An illustrative structure that may be used for the antifuse transistor 58 is shown in
As shown in
To program the antifuse, it is necessary to raise the voltage of the substrate VSUB to about 0.7-1.0 volts (whether using an avalanche-breakdown antifuse or a Zener-breakdown antifuse). When the entire substrate of the integrated circuit is electrically interconnected, capacitive loading effects tend to resist efforts to raise the substrate voltage VSUB. Accordingly, a deep n-well isolation region 140 may be used to electrically isolate p-type substrate region 88 a from the rest of p-type substrate 88 b. As shown by dotted line 142 and n-well doping region 140′, the n-type doping of the deep n-well isolation region may be extended further along the surface of the structure to widen the area of isolation.
When an n-well arrangement of the type shown in
The behavior of avalanche-breakdown antifuse transistors has been characterized using antifuse transistors of different sizes without deep n-well isolation region 140.
Transistors with lengths of 10 microns and widths of 0.15 microns have be observed to have off (unprogrammed) resistances of about 110 MΩ at 25C, unprogrammed resistances of about 23 MΩ at 85C, and programmed resistances of less than 1000Ω. The programming current for a 4.5 V programming pulse was about 33 μA.
Transistors with lengths of 2 microns and widths of 0.15 microns have been observed to have unprogrammed resistances of 550 MΩ at 25C and 110 MΩ at 85C. Resistances of less than 1000Ω have been observed for these transistors following programming. The programming current for a 4.5 V programming pulse was about 13 μA.
As seen, the unprogrammed (off) resistances of these illustrative antifuse transistors is more than four orders of magnitude greater than the programmed resistances of the transistors. This provides good resolution between the high and low states at the output VOUT. The wide range between the programmed and unprogrammed resistance values also provides a large design window for the series resistor R1. As power supply voltages scale with future improvements with process technology, the change in the unprogrammed resistance should not be large. In addition, the width of the antifuse transistor may be scaled down to maintain large enough unprogrammed resistances. Further improvements in antifuse performance (e.g., reduced programming voltage) may be obtained by using the deep n-well isolation region 140.
The programming power of the illustrative antifuses has been observed to be lower than 150 μW. Further reductions in the antifuse transistor width will result in lower programming powers.
The programming time may be 10 ms-50 ms or less. This short programming time allows many antifuses to be programmed in a relatively short time. Moreover, multiple antifuses may be programmed in parallel to further reduce the total programming time for the programming operation.
The current-voltage behavior of a two-diode Zener-diode stack has been observed, as shown in the current versus voltage plots of
Antifuse 58 can be programmed electrically at low powers. Programming powers may be 10 to 100 times lower than other types of antifuses. The lower powers reduce the risk of thermo-mechanical damage to process layers on the circuit. The low powers and electrical programming capabilities of the antifuse also reduce equipment overhead requirements and reduce programming times.
The foregoing is merely illustrative of the principles of this invention and various modifications, can be made by those skilled in the art without departing from the scope and spirit of the invention.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4207556||Dec 5, 1977||Jun 10, 1980||Nippon Telegraph And Telephone Public Corporation||Programmable logic array arrangement|
|US4433331||Dec 14, 1981||Feb 21, 1984||Bell Telephone Laboratories, Incorporated||Programmable logic array interconnection matrix|
|US4609986||Jun 14, 1984||Sep 2, 1986||Altera Corporation||Programmable logic array device using EPROM technology|
|US4617479||May 3, 1984||Oct 14, 1986||Altera Corporation||Programmable logic array device using EPROM technology|
|US4642487||Sep 26, 1984||Feb 10, 1987||Xilinx, Inc.||Special interconnect for configurable logic array|
|US4761768||Mar 4, 1985||Aug 2, 1988||Lattice Semiconductor Corporation||Programmable logic device|
|US5070384||Apr 12, 1990||Dec 3, 1991||Actel Corporation||Electrically programmable antifuse element incorporating a dielectric and amorphous silicon interlayer|
|US5291434||Jul 11, 1990||Mar 1, 1994||Gemplus Card International||MOS fuse with oxide breakdown and application thereof to memory cards|
|US5303199||Feb 19, 1991||Apr 12, 1994||Sharp Kabushiki Kaisha||Redundant memory device having a memory cell and electrically breakable circuit having the same dielectric film|
|US5463244||May 26, 1994||Oct 31, 1995||Symetrix Corporation||Antifuse programmable element using ferroelectric material|
|US5510730||Jun 21, 1995||Apr 23, 1996||Actel Corporation||Reconfigurable programmable interconnect architecture|
|US5646879 *||Dec 1, 1994||Jul 8, 1997||Micron Technology, Inc.||Zener programmable read only memory|
|US5774011||May 28, 1997||Jun 30, 1998||International Business Machines Corporation||Antifuse circuit using standard MOSFET devices|
|US5831923||Aug 1, 1996||Nov 3, 1998||Micron Technology, Inc.||Antifuse detect circuit|
|US5844298||Mar 28, 1997||Dec 1, 1998||Micron Technology, Inc.||Method and apparatus for programming anti-fuses|
|US6108261||Jun 29, 1999||Aug 22, 2000||Hyundai Electronics Industries Co., Ltd.||Repair circuit for redundancy circuit with anti-fuse|
|US6125069||Oct 19, 1999||Sep 26, 2000||Nec Corporation||Semiconductor memory device with redundancy circuit having a reference resistance|
|US6240033||Jan 10, 2000||May 29, 2001||Hyundai Electronics Industries Co., Ltd.||Antifuse circuitry for post-package DRAM repair|
|US6456546||Dec 18, 2000||Sep 24, 2002||Hyundai Electronics Industries Co., Ltd.||Repair circuit using antifuse|
|US6477094||Dec 18, 2000||Nov 5, 2002||Hyundai Electronics Industries Co., Ltd.||Memory repair circuit using antifuse of MOS structure|
|US6674667||Feb 13, 2001||Jan 6, 2004||Micron Technology, Inc.||Programmable fuse and antifuse and method therefor|
|US20040100849||Nov 27, 2002||May 27, 2004||David Novosel||Method of utilizing a plurality of voltage pulses to program non-volatile memory elements and related embedded memories|
|1||"Analyzing the process window for laser copper-link processing" Solid State Technology-Semiconductor manufacturing and wafer fabrication, Jan. 8, 2003.|
|2||Alexander Kalnitsky et al. "CoSi2 integrated fuses on poly silicon for low voltage 0.18 um CMOS applications" (C) 1999 IEEE.|
|3||JB Bernstein et al. "Analysis of laser metal-cut energy process window" IEEE Trans. on Semicond. Manuf., pp. 228-234, May 2000, vol. 13, Issue 2.|
|4||JB Bernstein et al. "Laser energy limitation for buried metal cuts" IEEE Electron Device Letters, pp. 4-6, Jan. 1998, vol. 19, Issue 1.|
|5||Jinbong Kim et al. "Three-Transistor One-Time Programmable (OTP) ROM Cell Array Using Standard CMOS Gate Oxide Antifuse", IEEE Electron Device Letters, vol. 24, No. 9, Sep. 2003, pp. 589-591.|
|6||Mohsen Alavi "A PROM Element Based on Salicide Agglomeration of Poly Fuses in a CMOS logic process" IEEE International Electron Devices Meeting, Dec. 1997.|
|7||Noriaki Sato et al. "A New Programmable Cell Utilizing Insulator Breakdown", IEDM 1985, pp. 639-642.|
|8||RT Smith et al. "Laser Programmable redundancy and yield improvement in a 64 K Dram", IEEE Journal of Solid-State Circuits, pp. 506-514, Oct. 1981, vol. 16, Issue 5.|
|9||U.S. Appl. No. 10/780,427, filed Feb. 17, 2004, Chih-Ching Shih et al.|
|10||V. Klee et al. "A 0. 13 um logic-based embedded DRAM technology with electrical fuses, Cu interconnect in SiLK, sub-7ns random access time and its extension to the 0.10 um generation," IEEE (c) 2001 IEDM (International Electron Devices Meeting), Technical Digest, pp. 18.5.1-18.5.4.|
|11||Wei Zhang et al. "Laser-Formed Vertical Metallic Link and Potential Implementation in Digital Logic Integration", PROC MAPLD, MD, 1999, pp. 1-7 (B5).|
|12||Wei Zhang et al., "Energy Effect of the Laser-Induced Vertical Metallic Link", IEEE Transactions on Semiconductor Manufacturing, vol. 14, No. 2, May 2001, pp. 163-169.|
|13||Wei Zhang et al., "Reliability of laser-induced metallic vertical links," IEEE Trans. on Advanced Packaging, pp. 614-619 Nov. 1999, vol. 22, Issue 4.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7679426 *||Jan 19, 2005||Mar 16, 2010||Hewlett-Packard Development Company, L.P.||Transistor antifuse device|
|US7884349||Sep 11, 2008||Feb 8, 2011||Unity Semiconductor Corporation||Selection device for re-writable memory|
|US7893520||May 12, 2008||Feb 22, 2011||International Business Machines Corporation||Efficient interconnect structure for electrical fuse applications|
|US7902611||Nov 27, 2007||Mar 8, 2011||Altera Corporation||Integrated circuit well isolation structures|
|US8027215||Sep 2, 2009||Sep 27, 2011||Unity Semiconductor Corporation||Array operation using a schottky diode as a non-ohmic isolation device|
|US8031509||Dec 18, 2009||Oct 4, 2011||Unity Semiconductor Corporation||Conductive metal oxide structures in non-volatile re-writable memory devices|
|US8133767||Dec 22, 2010||Mar 13, 2012||International Business Machines Corporation||Efficient interconnect structure for electrical fuse applications|
|US8194431||Oct 9, 2010||Jun 5, 2012||Silicon Line Gmbh||Programmable antifuse transistor and method for programming thereof|
|US8258813||Jan 11, 2010||Sep 4, 2012||Silicon Line Gmbh||Circuit and method for driving at least one differential line|
|US8525435||Nov 19, 2010||Sep 3, 2013||Silicon Line Gmbh||Circuit arrangement and method for controlling light emitting components|
|US8565039||Jul 23, 2012||Oct 22, 2013||Unity Semiconductor Corporation||Array operation using a schottky diode as a non-ohmic selection device|
|US8609534||Sep 27, 2010||Dec 17, 2013||International Business Machines Corporation||Electrical fuse structure and method of fabricating same|
|US8630108||Mar 31, 2011||Jan 14, 2014||International Business Machines Corporation||MOSFET fuse and array element|
|US8824898||Apr 8, 2011||Sep 2, 2014||Silicon Line Gmbh||Circuit arrangement and method for transmitting TMDS encoded signals|
|US8848425||Sep 10, 2013||Sep 30, 2014||Unity Semiconductor Corporation||Conductive metal oxide structures in non volatile re-writable memory devices|
|US8855154||Dec 22, 2009||Oct 7, 2014||Silicon Line Gmbh||Circuit and method for controlling light-emitting components|
|US8912627||Feb 26, 2013||Dec 16, 2014||International Business Machines Corporation||Electrical fuse structure and method of fabricating same|
|US8941110||Nov 17, 2011||Jan 27, 2015||International Business Machines Corporation||E-fuses containing at least one underlying tungsten contact for programming|
|US9107261||Dec 22, 2009||Aug 11, 2015||Silicon Line Gmbh||Circuit and method for controlling light-emitting components|
|US9293702||Sep 3, 2014||Mar 22, 2016||Unity Semiconductor Corporation||Conductive metal oxide structures in non-volatile re-writable memory devices|
|US9385025||Dec 17, 2014||Jul 5, 2016||Globalfoundries Inc.||E-fuses containing at least one underlying tungsten contact for programming|
|US9484254||Nov 1, 2013||Nov 1, 2016||International Business Machines Corporation||Size-filtered multimetal structures|
|US9548270||Sep 15, 2015||Jan 17, 2017||International Business Machines Corporation||Electrical fuse with metal line migration|
|US20060160318 *||Jan 19, 2005||Jul 20, 2006||Schulte Donald W||Transistor antifuse device|
|US20090016094 *||Sep 11, 2008||Jan 15, 2009||Unity Semiconductor Corporation||Selection device for Re-Writable memory|
|US20090278229 *||May 12, 2008||Nov 12, 2009||International Business Machines Corporation||Efficient interconnect structure for electrical fuse applications|
|US20100109794 *||Jan 11, 2010||May 6, 2010||Martin Groepl||Circuit and method for driving at least one differential line|
|US20100157658 *||Dec 18, 2009||Jun 24, 2010||Unity Semiconductor Corporation||Conductive metal oxide structures in non-volatile re-writable memory devices|
|US20100172384 *||Dec 22, 2009||Jul 8, 2010||Martin Groepl||Circuit and method for controlling light-emitting components|
|US20100172385 *||Dec 22, 2009||Jul 8, 2010||Martin Groepl||Circuit and method for controlling light-emitting components|
|US20110080765 *||Oct 9, 2010||Apr 7, 2011||Silicon Line Gmbh||Programmable antifuse transistor and method for programming thereof|
|US20110092031 *||Dec 22, 2010||Apr 21, 2011||International Business Machines Corporation||Efficient interconnect structure for electrical fuse applications|
|US20110121742 *||Nov 19, 2010||May 26, 2011||Silicon Line Gmbh||Circuit arrangement and method for controlling light emitting components|
|U.S. Classification||365/225.7, 438/600, 257/E27.102, 438/983, 257/E21.666, 438/467, 257/E27.112, 257/106, 438/131, 257/E23.147, 257/530, 365/96, 257/529|
|International Classification||H01L23/525, H01L21/8246, H01L27/112, G11C17/16, H01L27/12, G11C17/18|
|Cooperative Classification||H01L2924/0002, Y10S438/983, H01L27/11206, H01L23/5252, H01L27/1203, G11C17/16, H01L27/112|
|European Classification||H01L27/112P, G11C17/16, H01L27/112, H01L23/525A|
|Feb 18, 2011||FPAY||Fee payment|
Year of fee payment: 4
|Feb 25, 2015||FPAY||Fee payment|
Year of fee payment: 8